This application is responsive to Program Announcement PAR-08-268, the AHRQ Small Research Grant to Improve Health Care Quality through Health Information Technology (HIT). This application addresses the "Research Area" of "Medication Management" evaluating errors at the point of delivering ambulatory prescriptions. This application meets the "Specific Purpose" of the FOA by "generating information necessary for future health IT implementation projects." Our research design includes one of the suggested approaches for research responsive to this FOA - "self-contained HIT research projects." The PI is has new-investigator status, in his final year of an AHRQ-sponsored career development award. The objective of this R03 study is to expand upon preliminary work that evaluated frequencies of Look-alike, Sound-alike (LASA) errors among 11 selected pairs of pediatric medications. It is necessary to identify a larger set of candidate LASA drug pairs for intervention and refine the method for "flagging" individual prescriptions as potential LASA errors. The proposed study will utilize both qualitative and quantitative methods in completing study aims. We will evaluate a list of >1500 LASA drug pairs, using a modified Delphi method to reduce this list of LASA drug pairs to a list of approximately 200 drug pairs felt to be problematic in ambulatory pediatric care. We will then estimate rates of LASA errors utilizing an algorithm developed previously. There are 2 aims to this study. Aim 1: Identify a subset pediatric-specific, problematic LASA drug pairs. Using a modified Delphi approach, we will reduce the list of 1500 known LASA pairs to a targeted list of approximately 200 LASA drug pairs that are used in ambulatory pediatric practice and felt to be of high risk of harm to a child should a substitution occur. Aim 2a: Estimate frequencies of potential LASA substitution errors in these drug pairs, using South Carolina Medicaid data from 2000-2008. These calculations will build upon pilot efforts evaluating a list of 11 drug pairs using a similar method, completed as part of the PI's career development award. Aim 2b: Determine the positive predictive value ("true error rate") of the screening alerts identified for Aim 2a. We expect that this course of inquiry will ultimately lead to a computerized set of pediatric-specific LASA screening alerts that could be implemented in the pharmacy setting to reduce LASA errors for children. This analytic approach will also have future applicability to inpatient pediatric medication safety, as a similar approach could identify a list of pediatric-specific inpatient drugs to incorporate into CPOE systems. This study will ultimately improve pediatric prescribing safety by developing an expanded panel of LASA drug pairs commonly used in pediatric ambulatory care and considered at high risk to patients should a substitution occur. Such a targeted, pediatric list does not currently exist. Work completed as part of this R03 application will lay the groundwork for development of a larger-scale implementation study in pharmacy settings, with the goal of reducing pediatric ambulatory LASA errors. PUBLIC HEALTH RELEVANCE: Project Narrative - Prevention of medication errors is a priority in US healthcare, and Health Information Technology (HIT) offers substantial promise to reduce those errors. This study will identify a candidate list of LASA drug pairs commonly used in pediatric care and will estimate frequencies of LASA errors in those drug pairs using approximately 18 million ambulatory prescriptions. The findings of this study will be used to further develop a real-time, pharmacy based screening program to identify potential LASA errors at the point of prescription delivery, thereby meeting the goals of the program announcement "to improve the quality, safety, effectiveness, and efficiency of health care through the implementation and use of health IT."